PID, Controlling Position with Velocity Clamping

[Taylor Turner]

I'm using a three term equation to control position, but the velocity profile from point A to B is failing expectation.


I'm assuming the position target would have its own pid with velocity as the process variable, but I wouldn't want the position target to decrease is value and move the opposite direction. That would look horrible.


On the opposite side of the equation the control variable can be clamped when the velocity is met or exceeded, but I would think my terms would run away and not be there when the CV needs to decrease value.


Is the answer more obvious?

 

[Peter Nachtwey]

Writing a good target generator is more difficult than writing a PID.
Basically, you are doing it wrong.
You should use a system of equations as a function of time.
You should have 5 equations and 5 unknowns. This assumes that the maximum velocity is reached. If not then there will only be a ramp up and ramp down so now there are only 4 equations and 4 unknowns.

 

[Taylor Turner]

Now it's getting fun.
I'll start with answering what the denominator should be in dx/dt, given the initial error.

 

[drbitboy]

Is the answer more obvious?

No more so than the question .